This is an old revision of the document!
The Return of Depression Economics by Paul Krugman, 1999. London: Penguin
By July 1997, twenty years of events had placed capitalism in a uniquely triumphant position in its history: starting with Chinese reforms in 1978, feeding off the Soviet collapse and buoyed by the seeming triumph of the emerging SE Asian economies. Self-congratulation at the flawless management of capitalist economies had returned – just as on the eve 1970s stagflation, economic journals were asking, 'Is the business cycle obsolete?' so too, in 1997, Foreign Affairs published an article titled 'The end of the business cycle?'
Premature, perhaps, but it is important to have a working model of why depressions happen. Krugman uses the Sweeneys' baby-sitting cooperative as his starting point, modelling a depression as a failure of demand, perhaps caused by a scarcity of money. The most important conclusion to be drawn from this model is: bad things happen to good economies – no inherent weakness in the baby-sitting economy lead to its failure of demand, and so it is not advisable to assert that simply because an economy is in good, robust shape, that it is now immune from depressions.
Before World War II the role of demand management simply was not understood; from the War to the end of the sixties, the principles of Keynesian demand management seemed to be successfully adopted. in the 1970s Vietnam and the oil shocks proved that not all economic circumstances were easily manageable but by the 1980s it seemed that for the most part such management – even if only through monetary mechanisms – was proving successful. In 1987 the stock market crashed, losing as much value in the crash's first day as was lost in the first day of 1929 – but the Fed responded quickly, pumping cash into the system and there was very little impact on the real economy. By 1997, then, it did appear that even if the business cycle was not eliminated it might at least have been decisively tamed.
The advent of information technology was clearly having an impact on at least the perception of capitalism, even if the real value for the economy was being overrated – for the first time in a generation, certain markets were being dominated by intrepid entrepreneurs rather than staid corporations, and the IT revolution had an enormously visible impact on the environment in which most in the West worked: the landscape of the typical office. And the first beneficiaries of globalisation – the SE Asian economies – were now reaping enough unquestionable success to provide amply ammunition to those who argued that a globalised, export-led development path was the decisive means to escape poverty. There were problems: Japan had still not recovered from a recession that had begun in the early nineties, Europe was finding it difficult to force unemployment down as far as it would like, particularly amongst its young and even during recoveries, critics of capitalism pointed to those remaining in extreme poverty in the US and the working conditions of the poorest industrial workers in the developing economies – and the vast swathes in Africa and South Asia who were still untouched by this new growth.
But the prospects for the world economy in general, and for capitalism in particular, seemed better than they had been in living memory, better than anyone could have imagined a decade or two earlier.
It is often tempting to try to use hindsight to arrange history in such a way that events ought to have been predictable beforehand – in this case, to find some economic weakness or instability that, if only somebody had looked hard enough, would have betrayed SE Asia's vulnerability to the crisis that hit in July 1997. But the truth is that Asia's economic miracle was real, and it thoroughly deserved the hype it received before the crash. The economic 'take-off', from an acutely impoverished rural economy to a steadily industrialising one, especially at the remarkable rates seen in SE Asia – compare 7 per cent in South Korea with 8 per cent in Japan post-war, 1.3 per cent in Britain and 2.2 per cent in the US – has proved so illusive in other areas of the world and at other periods and its importance cannot easily be overstated. That these countries include about a quarter of the world's population makes the achievement especially significant.
However, one characteristic of Asian growth is interesting – total factor productivity (TFP, the amount of output that can be generated per unit of input, whether that input is capital or labour) did not rise, in many economies it was declining gently. In the industrialised countries, growth is almost entirely a consequence of increases in TFP, whereas in Asia, like in the Soviet Union, growth was a consequence of factor mobilisation and capital accumulation. Consequently, growth by this model could not continue indefinitely – eventually capital accumulation has to level off at some proportion of output and at that point growth is only possible through improvements in TFP.
Whether an 'Asian system' really existed is a moot point. These economies, although highly diverse in ethnicity, religion and other social factors, nevertheless displayed economic similarities, particularly high savings rates, and very good basic education leading to high literacy and numeracy – and this accumulation of physical and human capital in turn formed the basis for their growth. But it is less clear whether other features of the 'Asian system' – particularly the close association of business and the state, in which the state protected new industries and provided subsidised credit and generally actively involved itself in economic development was crucial to the Asian modality of economic development or not. This relationship between the state and private sector was given a lot of credit for the successes by some observers, and much of the blame for the crash by others, when it would come to be dubbed 'crony capitalism'. We should be sceptical of both sets of claims.
From the Mexican Revolution beginning in 1911 until the late 1970s, the Mexican economy had a fairly stable form: inward looking, severe tariffs and quotas, low growth but very stable. The system had been implemented by a revolutionary socialist government hostile to their internationalist free-market predecessors, but quickly became entrenched between a private oligarchy favoured by protected markets and control of import quotas, a formally democratic but never-actually-contested government supported by the oligarchs, and a kind of labour aristocracy in the protected industries. In the late 1970s, the stability of governance began to break down – in the wake of the oil price hikes (Mexico being an exporter) and new oil discoveries, government spending snowballed, still managing to take on new foreign debt, opening itself up to the wrong side of the crisis in 1982. After a 1981 peak growth collapsed, through the 1980s failing to keep up with population growth. In the mid-1980s the government progressively favoured reform, impressed by Chile's performance seemingly on the back of market reforms, removing the enormous protection of local firms and moving towards the eventual signing of NAFTA in the early 1990s, buoyed by the Brady Plan to restructure Mexico's debt in 1989. The US' liberal treatment of Mexico in part signalled the real fear that America felt of a more radical socialist government taking power on its southern border, not an unthinkable prospect with Mexico's history and the debt crisis of the early 1980s. There was extreme international confidence in Mexico by 1993, but there followed a series of fairly minor policy mistakes. Growth remained sluggish, up from 1.3 per cent from 1981-89 to only 2.8 per cent 1990-94 and despite a peg to the dollar, relatively high inflation continued, raising Mexican prices sufficiently that many warned that the peso was overvalued. In 1994 the Zapatista rebellion began and the likely next-president was assassinated. Government reserves were run down in the run-up to an election and a devaluation seemed more or less inevitable. Unfortunately, it was badly handled. The two rules of successful devaluations are
Both rules were broken: the peso was devalued 15 per cent, only half of the adjustment most observers called for, and anti-investor rhetoric did not reassure the investor community. The government was ultimately unable to support the peg, and the peso ultimately lost half of its value. Interest rates were soaring and government debt – dollarised as a means of restoring confidence in the peso – was inflated by the devaluation and had a serious impact on the budget. Against Congress' wishes, the Treasury found an obscure fund designed as a means of supporting the dollar from which it withdrew $50 billion to support the peso and, with a lot of luck, this was sufficient to bring the crisis under control.
Argentina had a more convoluted history. Benefiting from a good British-build and funded rail network, before the First World War it was a rich economy built on agriculture and beef exports, but as with other export-oriented economies it was hit hard by the Depression. It adopted a heterodox policy response faster than most, though, and printed its way out of crisis more rapidly than Europe and the US. Unfortunately, the legacy of heterodox successes before the Second World War left a legacy of rather inconsistent monetary policy and recurrent inflations. The debt crisis hit hard in 1982 at the same time as the Falklands debacle, and an attempt to institute a new, more stable currency ended in a disastrous hyperinflation of 3000 per cent. In 1989 a reformist government took office and made substantial changes. The always-successful beef and agriculture sectors had long effectively subsidised the rest of a hideously inefficient economy; tariffs were slashed. The government attempted to permanently stabilise the currency by resurrecting the traditional colonial concept of a currency board – a particularly hard form of peg in which every peso in circulation was backed by government dollar reserves, with a government guarantee to support an exchange rate of 1:1. Output leapt by 25 per cent over the next three years.
The Mexican crisis had a perhaps irrational knock-on effect on Argentina. Foreign debts were called in – perhaps by American banks that couldn't be bothered to distinguish between one Latin American company and another. The currency board system, highly robust to speculative attack suddenly revealed another kind of weakness – that the central bank could not extend the money supply without increasing its foreign reserves. The calling in of foreign loans forced the Argentine money supply to contract instead – and there was nothing the government or banking system could do to prevent it. Credit for businesses was insufficient and bankruptcies ensued. Lacking faith in the banking system, depositors rushed to withdraw and dollarise their accounts. The currency board system effectively made it impossible for the government/central bank to act as the lender of last resort – it was no more able to print pesos than it could print gold, at least without abandoning the currency board. It desperately needed dollars, which came, luckily enough, in the form of $12Â billion from the World Bank. The contraction in both economies was severe – bigger than that in the first year of the 1982 crisis, but the rescues worked and the crisis was surprisingly short, and both economies recovered their strength quickly.
Unfortunately, the lessons drawn from the experience were the wrong ones. It was supposed that this was a particularly Latin American crisis, a consequence of the peculiar Mexican system, and there were no wider implications. Moreover, the American rescue had been a meteoric success, supposedly proving the renewed insight and power of clever economists to control whatever crises could now be thrown at them.
These were the wrong lessons. It was not appreciated that the errors in Mexican policy-making, although real, were out of all proportion to the scale of the crisis – it seemed like an exceptionally severe punishment for such misdemeanours. Moreover, the rescue had been achieved primarily by a sleight of hand by the Treasury over Congress' disapproval – hardly an indication that a robust international system was working. And in the Mexican case the political relationship was easily managed – the US was particularly anxious to assist an economy which is too close to the US for America to see fail, and Mexican politicians were not above asking for help from Washington and accepting a little policy direction. Perhaps above all, there was a failure to understand just how lucky the rescue package had been in the effect that it had had – in truth there was little genuine reason to believe that intelligent policy intervention had been a more decisive factor than dumb luck in getting things back on track.
Japan's economy was very successful in the period 1953-73, growing at around 9 per cent per annum, the most rapid sustained industrial growth that had then been recorded, even faster than the early years of Stalinism. This success didn't really gain attention from the West until it was already decaying; from 1973 onwards growth slowed suddenly – at the same time as it did throughout much of the world – to around 4 per cent. Although debate on Japan's success was not particularly timely, it was extensive. Opinion divided into two main camps:
Advocates of the latter stressed the unusual features of the Japanese economy, particularly MITI and the Finance Ministry, who were heavily (but decreasingly) involved in directing investment and economic activity. The government strategically selected industries for the economy to focus on, the industry would be temporarily protected from imports, then there would come a phase of export penetration, in which firms would gain market share abroad at the expense of profits, and then once established, focus would shift to the next industry. This was particularly successful in the automotive and consumer electronics industries. One reason this was possible was the insulation of Japanese firms from short-term financial pressures – several Japanese firms tend to be organised around a bank, these firms own shares in one another and receive their finance from the bank rather than through bond or share issues. They are therefore insulated from market pressures to be short-term profitable – they can play the long game.
Problems began in the late 1980s. Banking deregulation had begun from around 1980, as it had in much of the world. Deregulation created a moral hazard problem: since it was widely recognised that the government would ultimately bail out a banking collapse, nobody in the private sector was responsible for the full risks of dangerous investment. As in other parts of the world, investors exploited the moral hazard, creating an unsustainable bubble. Unlike other areas of the world, Japan managed this bubble without a sudden crash, but its gradual deflation still lead to a recession. There followed the usual array of depression-fighting tactics during the course of the 1990s – but none of them worked:
There are two objections to the expected-inflation solution, one good and one bad:
The Asian financial crisis began in Thailand. Thailand has a larger population than the UK, but is one of the smaller Southeast Asian economies. It took off relatively late, industrialisation beginning in earnest around 1980, but its industrialisation was successful, growing at 8 per cent or more per year. Until 1990 or so, this investment primarily came from within Thailand – large export-oriented enterprises were financed from abroad, but smaller firms were created with individuals' savings and office blocks were built with bank loans from Thai deposits. After 1990 the capital market was increasingly liberalised, and a much larger volume of investment capital poured in.
In 1990 private capital flows to developing countries were $42 billion; official agencies like the IMF and the WB were actually financing more investment in the Third World than all private investors combined. By 1997, however, while the flow of official money had actually slowed, private flows had quintupled, to $256 billion. At first most of it went to Latin America, especially Mexico, but after 1994 it increasingly went to the apparently safer economies of Southeast Asia.
The Thai currency (the baht) was pegged to the dollar to provide predictability to business and confidence to investors. Whilst this capital was flowing in, the government was rapidly accumulating foreign reserves, and the money supply was expanding. The government tried to limit this expansion by issuing bonds – borrowing back the money they were printing to pay for yen – but this merely forced up the domestic interest rate, making the baht an even more attractive prospect. The uncontrolled monetary expansion fuelled not only a boom but a speculative bubble, largely in real estate. This in turn was further exacerbated by a moral hazard problem – the investment funds that were channelling Japanese and European investment into Thai firms were nepotisticly linked to government, creating the (realistic) expectation that they would be bailed out if they failed – providing an investment to put money in high-risk speculative investments rather than more conservative options.
During 1996-97 the boom lulled, and capital began to flow out of the country. The government faced the classic textbook dilemma of an approaching speculative attack. Its reserves were being run down and it could not maintain the exchange rate indefinitely. Its two options were:
The game provided a no-risk opportunity for investors, so the government required decisive and early action – in the event, it dithered, allowing its reserves to be run down. The result was inevitable and the crash came on 2nd July, 1997.
But it was much larger than expected. A devaluation of 15 per cent was anticipated – the true devaluation was more like 50 per cent. It turned out that the Thai economy, just like the other Southeast Asian economies, was particularly susceptible to the feedback mechanism of financial panic – even more so since capital market liberalisation had removed what damping effects had previously existed. Loss of confidence lead to a decline in the currency's value, rising interest rates and a slumping economy, which in turn lead to financial problems for firms, households and banks, which in turn further reduced confidence in the economy.
The final unexpected element was contagion. This appeared not to be based on economic fundamentals: the Southeast Asian economies were diverse in 1997, with different problems. Indonesia had a far smaller trade deficit and appeared to be in a much stronger position than Thailand at the onset of the crisis. These economies had real but insignificant trade and direct financial links. Probably the decisive mechanism for contagion was the degree to which the economies were lumped together in the minds of foreign investors. Much of the money that had been pumped into these economies was placed in generic 'Southeast Asian' funds which did not distinguish one from the other. As Thailand failed, so money was withdrawn from these funds and the average investor, not understanding the regional differences clearly, began to wonder whether Thai weakness indicated that other Southeast Asian economies might suffer similar fates. The loss of confidence became a self-fulfilling prophecy amongst several economies all of whom had the same tendency toward escalating financial panic.
There are three main things that a government or central bank might want to achieve using its exchange rate regime:
There are essentially three different forms of currency regimes. By a cute act of symmetry, each of the three makes two of these goals possible but prevents the other. Thus the choice of exchange regime is a choice as to which option is least painful for a government to give up. The possible regimes are:
Southeast Asia, prior to the crisis, had fixed rates – essentially regime (2). However, as a US official put it, “for developing countries, there are no small devaluations.” This is because of the combination of the confidence game and the double standard.
Although Southeast Asia was using an adjustable peg, and although this classically leaves a country vulnerable to speculative attack, most Southeast Asian countries did not have controls on capital flight. Instead, they were committed to retaining investor confidence – always prepared to do whatever necessary to convince speculators that their currency would not be devalued. There are two possible means of doing this:
In practice, (1) is only possible if speculators can be persuaded that things are improving somehow – otherwise capital simply continues to flow out and sooner or later the government runs out of reserves or credit. However, (2) is economically suicidal – to hike interest rates at the beginning of an economic downturn is guaranteed to seriously worsen the economic situation.
Instead of focusing on economic fundamentals, government policy is instead dictated by 'investor confidence'. The Keynesian Compact – the basic mechanism of sound economic management that has prevailed in the West since the Keynesian solution to the Great Depression – is thrown out in favour of anti-Keynesian policies intent on satisfying the irrational fear of capital markets: hiking interest rates to tempt capital back to the currency zone, and often balancing budget deficits 'in case' investors are concerned that these might be a sign of a weak economy.
This game is easy to win in the West. In the aftermath of the Southeast Asian crisis, the Australian dollar was allowed to lose 25 per cent of its value. But then the slide stopped – the economy was judged fundamentally sound and investment poured back in, recognising that the Australian economy now represented a sound investment after such a devaluation. Such confidence is not extended to Third World economies, and the reason is essentially investor prejudice.
In the Southeast Asian crisis, the IMF made reducing budget deficits a condition of loans, and most of the economies adopted anti-Keynesian policies (excessive interest rates, increased tax, lowered spending) in a bid to win (an irrational) 'investor confidence'. Reduced budget deficits were clearly a mistake – these were not a initially a concern to investors, but where they were enacted they further weakened these economies and where they were not enacted, the resulting IMF scorn further fuelled the panic. And the IMF also insisted on structural reform – which was entirely inappropriate as such reform was not relevant to the issue in hand and the money was needed quickly.
However, the real issue was over interest rates. The IMF advised these countries to raise them to extremely high levels, which clearly damaged the domestic economy. Some have argued – notably Jeffrey Sachs – that instead governments should instead have allowed their currencies to float early in the crisis, predicting that the same thing would have happened as happened in Australia – a devaluation followed by a stabilisation. The IMF response is that Southeast Asia is not Australia and fundamental confidence in these economies would not have limited the extent of the devaluation. Instead, the result would have been a hyperdevaluation which would have caused not only the same financial distress (because of the increase in the cost of all the dollar-denominated debt that Southeast Asian firms had) and also high inflation because of the hike in the cost of all imported goods. We can't know for certain that Sachs is wrong on this one, but it's not obvious that he is right. The final option would have been to defend the rates to infinity without raising interest rates – which some have advocated – but in the absence of any change in government policy, it seems inevitable that the capital flight would simply have continued indefinitely and the same crisis would have happened anyway.
“Hedge funds don't hedge. In fact, they do more or less the opposite.” To hedge means to reduce risk – hedge funds' stock in trade is high-risk low-liquidity investment – areas in which they can earn a premium because more risk-averse agents won't touch them. They use investment capital from a small number of very rich individuals and institutions and leverage this capital very highly – sometimes by as much as a hundred times – so that they can make an enormous return from a slight fluctuation in asset, commodity or forex prices.
The first of the modern spate of speculative attacks was lead by George Soros against the pound sterling in 1992 – then part of the ERM and therefore subject to a government commitment to maintain the exchange rate within an agreed band. In practice, European monetary policy was set unilaterally by the Bundesbank, who had raised interest rates in order to prevent inflation that might otherwise have been caused by the enormous budget deficits it was running due to ongoing reunification. Britain was in recession and had entered the ERM with an overvalued currency. Low interest rates meant pressure on the pound.
Soros' Quantum Fund began by surreptitiously borrowing $15 billion in sterling and quietly exchanging for dollars. Towards the end of this conversion, he stepped up the publicity, giving press interviews exclaiming that the pound would be devalued, noisily converting the last of his pounds for dollars. The UK briefly hiked interest rates but were forced to back down against domestic pressure and after using $50 billion to defend the pound, allowed it to float. It lost 15 per cent of its value and stabilised – probably more or less what the UK needed.
Hong Kong initially survived the worst of the Southeast Asian crisis. In many ways it was the soundest economy in the region, with less corruption and more capital freedom – it had the benefit of a lot of investor confidence. It also had a currency board system, in which the Honk Kong Monetary Authority (HKMA) held a US dollar for every 7.8 HK dollars in circulation – and thus able to defend its currency to the hilt. However, by the middle of 1998 amidst a sea of depressed economies, the hedge funds conspired to attack the HK dollar. They borrowed stock from the HK exchange, sold it and converted the proceeds to US$. They were making a double bet: either that the currency would be devalued, giving them a return on the HK$-US$ conversion, or else the HKMA would hike interest rates to defend the HK$, driving down stock prices and giving the hedge funds a return on their short position in stock. It seemed like a risk-free bet.
But the HKMA fought back. Not only was its currency entirely backed by dollars, but it had considerable surplus dollar reserves on top. It took the unconventional – some would say heretical – step of frantically purchasing HK stock. It continually replaced the money that the hedge funds were taking out. It came under global criticism for this egregious affront to free-market principles (obviously, a few rich speculators have the divine right to manipulate a market – but the idea of a representative government doing the same in the public interest is an outrage) and the hedge funds rolled over their bets, expecting it to cave to international pressure. But the HKMA held out long enough for attention to be diverted by the unfolding Russian crisis, at which point the hedge funds had to call in their bets.
The hedge funds' attitude to Russia was different. In this case, a default was widely feared, but the hedge funds believed it would not be allowed to happen – primarily because of the American geopolitical concern that forcing Russia into undue financial chaos would create risks for the safe handling of the Russian nuclear arsenal. So the hedge funds predicted that the US (or the IMF at Treasury behest) would come to the rescue, again and again, with the necessary bailouts. They therefore sought to take advantage of the exorbitant Russian interest rates (up to 150 per cent) whilst other investors were fleeing. They took long positions in Russian government debt and other high-risk, illiquid assets and short positions in much safer, more liquid assets.
Then Russia defaulted. Those who were owed assets by the hedge funds called them in. The hedge funds, desperately needing cash to supply their creditors, tried to sell their other long positions. But these markets – for high-risk, illiquid assets – were dominated by the large hedge funds and were very shallow. Realistically, the only agents they could sell to were other hedge funds, but they too had lost money on the Russian default and if anything were looking to sell. These markets crashed: the most leveraged hedge funds took the rest down with them.
The recovery is difficult to explain. The Fed arranged for a buyout of one of the largest hedge funds, Long-Term Capital Management (LTCM), based in Connecticut. It also devolved unusual powers to Greenspan to cut interest rates. But it is difficult to understand how cutting interest rates could affect the fundamentals – if nobody can borrow then the rate at which they would if they could is irrelevant. Rather, the explanation seems to be that somehow, Greenspan managed to reassure the market enough to stave off a self-sustaining panic. Back to market neuroses trumping economic fundamentals…
Despite criticism from the IMF for its tight capital controls, even the IMF was glad of them when China provided the stable backbone that prevented the Asian crisis from becoming a total nightmare. It was therefore understandable that Malaysia announced the introduction of similar controls in the summer of 1998. What is more remarkable is that the Asian economies stuck with prevailing orthodoxy for so long. Incidentally, the week before they were announced, they had been advocated by Krugman.
But the timing was bad. By mid-1998 the crisis had more or less run its course, and the Southeast Asian economies were starting to recover naturally, particularly Korea. Additionally, a rather tenuous effect of the Russian crisis was helping to get things back on track – most of the hedge funds involved in Russia were short in yen to take advantage of low interest rates and a currency that seemed unlikely to appreciate, but the crisis forced them to repay this debt, pumping foreign currency into yen. Once more, there was a feedback effect: their injection was so large as to appreciate the yen, worsening their liabilities and forcing them to repay more of their debts, requiring further purchase of yen. The appreciating yen pulled up other Asian currencies, reducing the burden of dollar-denominated debt throughout the region.
The Russian crisis spread to Brazil – a country with little fundamental similarity, but which nevertheless proved vulnerable to panic. After a near hyperinflation in the mid-1990s, Brazil had introduced a new currency, the real, under a crawling peg regime. After the Russian default, the real came under attack, and Brazil raised interest rates and cut tried to cut budget deficits in order to stem the attack. But the austerity measures were too harsh to be workable, and one of Brazil's regional governments declared a default on debt it owed to government, sparking greater panic. Brazil devalued, but made the common mistake of not devaluing far enough – convincing the market that it is no longer committed to the peg whilst failing to remove the temptation to speculate against further devaluation. Two days later it allowed the real to float, but the consequences weren't as bad as predicted – the real fell only ten per cent and this was accompanied by a 33 per cent leap in the Brazilian stock market, which clearly believed that interest rates could now be permitted to fall. Brazilian firms were not as debt-laden as those in Southeast Asia, and things were looking brighter. Unfortunately, there followed a meeting in Washington. Ritual interest rate hikes were ordered, in order to consolidate defence of the currency – even though it seemed clear that the market both expected and wanted rates to be lowered. The real went into freefall.
[I]f the intention [of the IMF, in repeating orthodox advice] was to vindicate the policies of the past, the effect was just the opposite. Indeed, Brazil's crisis had the feel of a recurrent nightmare: once again, as in Mexico, Thailand, Indonesia, and Korea, a seemingly successful economy had gone to Washington in its hour of need, tried its best to follow the plan Washington devised, and been rewarded with a catastrophe...it seems that Brazil is now set for a crisis at least as bad as Thailand's or Korea's. --p150-1
'Depression economics' is, broadly speaking, the analysis of situations in which there is a failure of demand. It has been firmly in the background of economics more or less since the end of the Great Depression, for two basic reasons.
The specific set of silly ideas that has laid claim to the name "supply-side economics" is a crank doctrine which would have little influence if it didn't appeal to the prejudices of editors and wealthy men...
and yet in a world in which the problems of demand failures have been decisively solved it is natural for focus to shift to problems with supply. But depression economics is back: the stories detailed in this book all involve both a failure of demand and 'free-lunch problems': problems in which, unlike most economic issues, there is no trade-off – there are universal gains to exploit if only the situation can be properly understood and the proper management approach adopted.
The true scarcity in [Keynes'] world -- and ours -- was therefore not of resources, or even of virtue, but of understanding.
Our failure to understand and find remedies to these problems threaten the gains that have been made over the past few decades in freeing markets and liberalising international markets. In a demand-deprived world, the clear arguments in favour of free markets suddenly begin to collapse – for instance
Right now a tariff would increase employment in Argentina, and to pretend otherwise is intellectually dishonest.
It is also true that whilst the economies affected are not perfect, this does not absolve the system from blame for their malaise – all economies are imperfect and the system needs to be robust enough to protect them from demand failures without them enacting perfect policy. It is also absurd to welcome demand failures as an opportunity for pressing greater reform and improvement of fundamentals – it is unnecessarily painful and besides, it is generally less painful to make improvements to a succeeding economy than to one with its back against the wall. Blaming the victims is not a solution, it is an excuse for not offering a solution.
In the rich world, the lesson of Japan is that we, as advanced economies, are no longer fighting a terrible inflationary threat as we were in the 1970s. Obsession with inflation is not only unnecessary, but becoming increasingly dangerous. The more immediate threat is a liquidity trap in which rates cannot be lowered sufficiently to stimulate recovery. The Western world should therefore take care to maintain inflation at or above 2 per cent during a boom, in order to provide a floor of -2 per cent to real interest rates. In Japan, planned inflation is the obvious and most effective solution to its impasse: the central bank should manage the money supply so as to set the long-run inflation rate at around 4 per cent.
The poor world's problems are much more difficult. But the standard response to speculative attacks needs to change – current IMF doctrine of trying half-heartedly to support the currency without any credible assurance that it is possible whilst destroying the domestic economy with punitive interest rates is not the answer. There is no consensus on what should replace this policy with critics of the IMF sharply divided, but the truth is that any of the alternative options would be an improvement on current practice – each of the three suggested responses has its benefits and could be valuable in the right circumstances. The overriding advice to policy-makers is do something and do it decisively – do anything but dither. The three main options are
 1953-1973, although Japan had widely industrialised before World War II, so this is a rather unfair comparison.
 I'm a bit unsure about this whole TFP thing. It seems to me, that beginning with an extremely capital-poor, rural economy, that even with very low technology a dollar of capital will create an enormous surge in output, whereas as capital formation progresses, the productivity of capital will gently decline. I would expect this process to offset the increasingly technologically efficient use of capital that would, in Western economies, show up as improvements in TFP. But maybe I Just Don't Understand.
 Whether it was an explicit conspiracy, an informal allusional agreement or a fortuitous coincidence hasn't been proven. Given the small number of players involved and their understanding of the game that they were playing, there hardly seems to be a practical difference.